The Optimization of the Multi-Atmospheric Ar-Xe Laser

Home , Stationary state

250 IEEE JOURNAL OF QUANTUM ELECTRONICS, VOL. 34, NO. 2, FEBRUARY 1998 The Optimization of the Multi-Atmospheric Ar–Xe Laser S. W. A. Gielkens, W. J. Witteman, V. N. Tskhai, and P. J. M. Peters

Abstract— The quasi-steady-state conditions of the multi- make this laser also a competitor for the well-developed CO atmospheric e-beam sustained Ar–Xe laser are investigated. It laser. The advantages of this new atomic laser are the absence is observed that the duration of the stationary period depends of dissociation and the regeneration of the laser gas, and on the e-beam current, discharge power deposition, and gas pressure. The laser efficiency can be as high as 8%. Beyond the the much shorter wavelength of about m compared to stationary period the efficiency drops. The pulse energy with CO lasers. optimum efficiency depends strongly on the gas pressure. The The full exploration of the new system requires a detailed maximum discharge efficiency of 5%–6% is at high pressure study of its parameters and its kinetic chain of the inversion not sensitive to the input power. The best results are obtained production. The understanding of the laser and the quantitative for 4 bar with a discharge input power of 8 MW/. The pulse duration with corresponding output energies is 12 "s with information on the kinetics can then be used for the design of 10 J/ and 16 "s with 16 J/ for e-beam currents of 0.4 and an optimized system. From an experimental point of view, it 0.9 A/cmPY respectively. An analysis of the quasi-steady-state is attractive to study this atomic xenon laser by means of an conditions that include the effects of electron collision mixing -beam sustained device because this technique has shown so and atomic quenching is presented. The effects of output power far to be most productive and efficient for radiation produc- saturation by the fractional ionization and atomic collisions are in agreement with the observations. The analysis clarifies the tion. Moreover, it allows to follow the effects of discharge optimum performance conditions. parameters and gas composition more or less independently and in this way unravels kinetic processes that otherwise may Index Terms— Electric discharge pumping, electron beam pumping, gas lasers, lasers, laser thermal factors, power lasers, be strongly mixed. For instance, the -beam produces a stable pulsed lasers. homogeneous plasma, independent on the gas pressure, so that the study of gas density effects is not hampered by plasma instabilities that in self-sustained discharges are automatically introduced by the increase of the gas pressure. In principle, I. INTRODUCTION the -beam sustained discharge allows to follow more or less HE STUDY of the laser transitions between the and independently the effects of discharge current, e-beam current, T bands of xenon [1]–[7] is of considerable interest for and gas pressure. several reasons. First of all, it is from a scientific point of In our previous work [7], we used a short e-beam pulse view remarkable that these lasing infrared transitions can be of only 1.2 s and a much longer discharge pulse. These very efficient up to 8% depending on discharge conditions experiments showed the fast drop of the output power after and that pulse energies up to 15 J with power densities of termination of the -beam. This work clarified the necessity several MW can be obtained [7]. Secondly, the apparent of simultaneous operation of discharge and -beam. Further- favorable kinetic chain of this laser process based on three- more, the experiments revealed the more or less quadratic body collisions challenged the development of CW systems dependence of the optimum input power on the gas pressure. with output powers in the order of watts [8], [9]. This became The experiments also brought forward the question to what successful with RF excitation of a mixture at 90 torr in narrow extent we are dealing with the quasi-steady state during the waveguide structures where output power densities of about simultaneous presence of the pulses and what the saturation 0.27 W/cm were obtained, which is two or three orders of mechanisms are. We particularly want to have more insight magnitude higher than what was previously known for low- into the quenching effects of electrons and atoms. To study pressure atomic discharge Xe lasers. This breakthrough in the these questions, we reconstructed our system to have simulta- gas laser development of obtaining high power combined with neous pulses for the -beam and sustainer of about 20 s. For the typical high optical quality opens the gate to many new this device, we observed the output waveforms as a function promising applications, e.g., the field of remote sensing and of -beam current, discharge current, and gas pressure. A communications. The high efficiency and high output power kinetic model is developed to get more insight into the kinetic processes. Manuscript received July 17, 1997; revised October 9, 1997. These investi- gations in the program of the Foundation for Fundamental Research on Matter were supported in part by the Netherlands Technology Foundation (STW). II. EXPERIMENTAL SETUP The authors are with the Department of Applied Physics, University of Twente, 7500 AE Enschede, The Netherlands. The electron gun is based on a plasma cathode and is Publisher Item Identifier S 0018-9197(98)01097-5. described elsewhere [10]. The -beam current density after

0018–9197/98$10.00  1998 IEEE GIELKENS et al.: OPTIMIZATION OF THE MULTI-ATMOSPHERIC Ar–Xe LASER 251 passing the 15- m-thick Ti foil was varied between 0.25 and 0.9 A/cm . In our experiments, the accelerator voltage was kept constant at 185 kV. The discharge circuit consists of three capacitors with a capacitance of 30 F each and two inductors each of 400 nH to provide for a more or less rectangular shape with a duration of about 20 s. The discharge is switched on by the -beam. The discharge is maintained between the foil and (a) an additional electrode. To avoid sputtering, we used the foil as anode. When this foil was used as cathode the sputtering resulted already in foil rupture at a current of 20 kA. The resonator consists of a flat totally reflecting Cu mirror and a plan-parallel ZnSe output coupler with a reflectance of 50%. These mirrors are separated by 90 cm. The distance between the electrodes is 2 cm and the cross section of the -beam is 3 53 cm . The laser extraction volume is 0.31 and the (b) base vacuum in the laser chamber was 5 10 bar. High- purity argon (99.9990%) and xenon (99.990%) were used. The beam and discharge current were measured by Rogowski coils. The accelerating and discharge voltage were measured by resistive voltage dividers. The contribution of an inductive element to the measured voltage appeared to be negligible. By multiplication of the measured discharge current and voltage the input power of the discharge was calculated. The power deposition by the -beam is calculated from stopping power (c) data [11]. The laser oscillates on several transitions between t the and levels of Xe. The temporal profile of the total Fig. 1. Temporal profiles of the (a) beam current density em, (b) discharge current sdis, and (c) laser output power out. output power is measured by a fast uncooled InAs photodiode (EG&G J12-18c) in combination with a CdTe window that transmits all laser lines but blocks visible radiation. The total product of the chosen -beam current and gas pressure because output energy is detected by a pyroelectric joulemeter (Gentec at higher input powers of the discharge when the stationary ED 500). By comparison of the measured energy with the period is short, it is observed that a substantial part of the measured waveform detected by the photodiode, the amplitude stationary output power is already present during the build-up of this diode signal is converted into units of power. time of the -beam. For each picture, the total input power can be inferred from Fig. 2. In this figure for each experimental condition the corresponding beam input power and discharge III. EXPERIMENTAL OBSERVATIONS power is plotted. The results of the stationary time as a function The typical behavior of the pulsed experiment is the ap- of the discharge power are shown in Fig. 3(a) and (b) for - pearance of the output pulse shortly after the onset of the beam current densities of 0.4 and 0.9 A/cm respectively. discharge pulse, followed by a quasi-steady-state regime where It is clearly seen that the stationary time strongly depends the -beam current, discharge current, and output power are on pressure. Although the experimental data are somewhat more or less constant and finally the region with the premature scattered owing to experimental fluctuations, the stationary fall-off of the output pulse whereas the discharge and -beam time is roughly inversely proportional to the discharge power. pulses are still present (see Fig. 1). The stationary period is Beyond this stationary regime, the output power and laser then determined by the time during which the output power efficiency decrease. The output power in the stationary regime is more or less constant. At the end of this period, we always as a function of discharge power is plotted in Fig. 4(a) and observe a continuous fast drop of the output. The present (b) for -beam current densities of 0.4 and 0.9 A/cm paper only considers the total lasing potential of the two bands respectively. The general behavior of increase of the output by investigation of the multiwave mode. We always observe power is, up to a value that depends on the gas pressure, that the total laser output of the oscillator with broad-band proportional to the discharge power. The higher the gas reflectors does not show any substantial modulation during pressure the larger the value of the discharge power that limits the stationary period whereas the observed individual lines this proportionality regime. The intrinsic efficiency of the laser are strongly modulated during this period which is due to with respect to the discharge power is plotted in Fig. 5(a) the well-known line competition. The experiments give us the and (b) for -beam current densities of 0.4 and 0.9 A/cm stationary duration of the output as a function of pressure, respectively. These values are calculated from the ratio of the discharge power, and -beam current. In the following, we will total output power minus the output power generated by the show various quantities as a function of the discharge power -beam only and the discharge power. The total efficiency is density for -beam current densities of 0.4 and 0.9 A/cm . determined by the ratio of the output power and total input The dissipated -beam power is not simply a constant times the power. When we plot the total efficiency versus the total input 252 IEEE JOURNAL OF QUANTUM ELECTRONICS, VOL. 34, NO. 2, FEBRUARY 1998

(a) (b) Fig. 2. For each experimental condition, the beam input power that corresponds to the given discharge power is plotted.

(a) (b) Fig. 3. The duration of the quasi-steady-state behavior of the laser as a function of the discharge power at e-beam current densities of (a) 0.4 and (b) 0.9 A/cmP.

power, as is done in Fig. 6(a) and (b) for -beam current [Ar] where is a constant and the beam densities of 0.4 and 0.9 A/cm we see that the maximum current density. These ions are lost by three-body collisions efficiency drops with total input power and that the highest with Ar to form Ar . The main process suffered by the efficiency of about 8% is reached for input powers below 10 molecular argon ions is the formation of ArXe in collisions MW at a low gas pressure. For each input power, the gas with Xe. The recombination of ArXe leads to the formation pressure can be optimized and the optimized pressure increases of the higher excited states of Xe, which are subsequently with input power. Finally we plotted the available output quenched by atomic collisions to reach the upper laser level energy per pulse during the stationary period as a function of manifold. The lower laser level manifold is at multi- discharge power and total power (see Fig. 7). It is remarkable atmospheric pressure mainly quenched by Ar to reach the that this output energy is sensitive to the -beam current and metastable level. There is also some radiative decay to that it has a maximum around 4 bar. the metastable level. Then, the metastable Xe atoms will produce ArXe excimers in three-body collisions with Ar. These excimers decay by radiative dissociation and form again IV. ANALYSIS OF THE SYSTEM the ground state. It is seen that the relevant xenon levels We shall discuss a kinetic model of the laser process by above the metastable levels are separated by about 1 eV, an means of a flow diagram of the kinetics shown in Fig. 8. energy comparable with the average electron energy of the It indicates the main species and kinetic reactions that are discharge. For that reason, it is generally accepted that the typically expected for the -beam sustained multi-atmospheric discharge mainly contributes to the excitation and ionization Ar–Xe laser with only 0.5% Xe. The electron beam mainly from the metastable level. Because of the low xenon content ionizes the argon gas proportional to its density [Ar] so the contribution of the -beam to the formation rate of Xe that the ion production by the -beam can be written as is negligible compared to the discharge contribution. In this GIELKENS et al.: OPTIMIZATION OF THE MULTI-ATMOSPHERIC Ar–Xe LASER 253

(a) (b) Fig. 4. The output power density during the stationary duration as a function of the discharge power for an e-beam with a current density of (a) 0.4 A/cmP and (b) 0.9 A/cmP.

(a) (b) Fig. 5. The intrinsic discharge efficiency versus the discharge input power density during the stationary time of the laser for (a) 0.4 A/cmP and (b) 0.9 A/cmP. This efficiency is defined by the ratio of the total output power minus the output power generated by e-beam pumping alone and the discharge power.

way, the discharge is effective in producing Xe which in kinetics of the laser inversion which depends strongly on the three-body collisions is converted into ArXe . In principle, discharge parameters and gas pressure, there is also consider- the kinetic chain of the discharge and laser process forms a able quenching of the inversion by collisional mixing of the closed cycle bounded by the metastable and ionization level of and manifolds by both electrons and atoms. For that Xe. In this steady-state process, the supply of metastable Xe reason, the laser performance is a strong interplay of discharge atoms by the -beam compensates for the above-mentioned power, -beam current, and gas pressure; each parameter can loss of metastable atoms that decay via the excimer to the be optimized in relation with the other ones. ground state. It is experimentally observed that after an early termination of the -beam the discharge impedance increases V. KINETICS drastically and the output drops. Below, we will describe a kinetic model for the steady-state The electrons are delivered by both the -beam and dis- behavior of the laser. The rate equation for Ar is mainly charge. The discharge conditions depend on the -beam current given by the following process: density. In the case of pure -beam pumping the average electron energy increases slightly with increasing -beam cur- Ar rent density. At lower -beam current and constant discharge Ar Ar Ar (1) power the reduced electric ﬁeld strength the drift velocity, and average energy of discharge electrons are higher where is a constant of proportionality, is the - and the electron density is lower. Apart from the formation beam current density, and is the formation constant of 254 IEEE JOURNAL OF QUANTUM ELECTRONICS, VOL. 34, NO. 2, FEBRUARY 1998

(a) (b) Fig. 6. The total efficiency as a function of the total input power density for (a) 0.4 A/cmP and (b) 0.9 A/cmP. This efficiency is defined by the ratio of the total output power and the total input power.

(a) (b) Fig. 7. The dependence of the output energy on the discharge power density during the stationary operation of the laser for (a) 0.4 A/cmP and (b) 0.9 A/cmP. Notice that this energy also peaks as a function of the pressure. molecular argon ions. The main process for Ar in our system rate proportional to the discharge power is described by (4) Ar Ar Ar Ar Xe (2) where is a constant of proportionality and is the rate where is the formation constant for molecular ArXe . The constant for formation of ArXe . The molecular ArXe molecular ArXe formation in the -beam chain is described formation in the discharge chain is then given by by ArXe Xe Ar ArXe (5) ArXe Ar Xe ArXe (3) The xenon metastables are produced by both the -beam and discharge. The production by the -beam is in the stationary state equal to the ionization rate of Ar. Similarly, in the where is the recombination rate constant of ArXe and discharge chain the production rate of the metastables is in our is the electron density. The main kinetic chain for Xe model equal to the ionization rate of Xe, which means that the is dominated by the discharge power density that ionizes discharge has no effect on the density of the metastables: the metastable Xe atoms, and by its quenching by three-body collisions to form ArXe . Since in the model the discharge Xe Ar Xe Ar (6) will ionize the metastable xenon atoms, we take the production GIELKENS et al.: OPTIMIZATION OF THE MULTI-ATMOSPHERIC Ar–Xe LASER 255

Fig. 9. Electron density versus the discharge power density at various pressures as calculated from (7).

Fig. 8. Scheme of the most important kinetic reactions and species for TABLE I the plasma chemistry of the Ar–Xe laser pumped by an e-beam sustained VALUES OF THE REACTION CONSTANTS OPTIMIZED TO MATCH MOST CLOSELY P discharge. The bold names are labels of the energy levels, whereas the italic THE EXPERIMENTAL RESULTS AT AN e-BEAM CURRENT DENSITY OF 0.9 A/cm ones are the particles involved in reactions indicated by a solid arrow. The dotted arrows denote radiative transitions. where is the rate constant for three-body quenching of the metastables. For the stationary state the electron density is equal to the sum of the ions, which can be deduced from (1)–(5) Ar Ar Xe Ar Ar (7)

The electron density is plotted in Fig. 9 as a function of the discharge power for various pressures. The values of the various constants are given in Table I. The values for and have been taken from [1]. The value for has been calculated from the beam input power and the by electron collisions. The electron collisions strive for ther- value for Ar [2]. The value of has been chosen such malization of the and manifolds characterized by the that the model ﬁts the laser output power as a function of the electron temperature. The quenching of the upper laser level discharge power (see the next section). It is seen that is not by the electrons is proportional to its density as well as the sensitive to the gas pressure. The metastable density becomes electron density according to (6) (9)

(8) where is a constant of proportionality. The quenching parameter depends on the plasma conditions like the average electron energy. At higher average energy the collision VI. LASER PROCESS frequency increases and consequently the quenching rate too. It In the following, we shall describe the main kinetics of the is expected that the higher the plasma conductivity determined laser process. The inversion is dominated on the one hand by the -beam current, the lower the average energy of by the excitation processes of the -beam and discharge. The the electrons and the smaller the quenching parameter. The production rate of the upper laser level with density will parameter will then depend on the -beam current density. In be in our model proportional to the ionization rate of argon order to get a reasonable ﬁt with the experimental observations, i.e. Ar . Since the ionization rate of xenon is taken we approximate with the relation . proportional to the discharge power, the production rate of Similarly the quenching by two- and three-body-collisions the upper laser level is in our model also proportional to of the atoms is given by the discharge power . On the other hand, the inversion is quenched by electrons and atoms. The lasing transitions (10) have high oscillator strengths so that they are tightly coupled 256 IEEE JOURNAL OF QUANTUM ELECTRONICS, VOL. 34, NO. 2, FEBRUARY 1998

Fig. 11. Plot of the optimal value of the output power density versus the pressure for the data of Fig. 12. Experimental values are indicated by symbols, Fig. 10. Curves of the output power as a function of the discharge power whereas the calculated points are depicted by the curve. at an e-beam current density of 0.9 A/cmP. For various pressures, these are calculated from (16) with values of the constants as mentioned in Table I. The symbols in this figure coincide with those in Fig. 4(b). The initial values for the parameters and are based on [2]. Together with and the final values of where and are proportionality constants and the these parameters are obtained by matching the calculated gas density practically equal to [Ar]. Including the stimulated curves with the measured data. These values are listed in emission, the upper laser level density is given by Table I. The output power according to (16) together with our experimental data are plotted in Fig. 10 for an -beam current density of 0.9 A/cm . In Fig. 11, the optimal output power is plotted versus the pressure for both the experimental data and (11) the calculated curves. These figures show that a reasonable where is the radiation density and the Einstein coefficient agreement with the experiments can be accomplished by for stimulated emission. Since the mentioned quenching pro- this model. Studying this result, we make the following cesses strive for thermalization between the laser levels, we conclusions. write for the lower laser level with density 1) For small input power of the sustainer, the output power scales proportionally. In this regime, is low enough for quenching to be negligible. In (16), this is expressed (12) by the term containing which is often referred to by fractional ionization. where is the pumping of the lower level and the 2) The output saturates with the discharge power which quenching of the lower level, which is proportional to the gas is due to the fact that increases with the discharge density. To maintain the inversion, is much larger than power, as seen in Fig. 9, and finally outweighs the and the decay of the lower laser level by three-body collisions pumping term. The calculated results are shown in is negligible. Fig. 10. The radiation production is equal to 3) The electron quenching depends on the fractional ionization . The electron density is not sensitive to the (13) gas density according to (7) and shown in Fig. 9. This The inversion is given by means that according to (16) the maximum obtainable power increases strongly with the gas pressure, which (14) is in agreement with the observations plotted in Fig. 4. It is also seen in Figs. 5 and 6 that the efficiency for where is the decay time of the resonator determined by the same input power at larger input powers increases its quality factor. For our system we find . With this with the gas pressure which can also be explained by we obtain by adding (11) and (12) in the stationary regime the decreasing fractional ionization with increasing gas (15) pressure. 4) From (16), it is predicted that atomic quenching is Substituting (15) into (11) and eliminating by negligible up to about 3 bar, whereas the strong electron (13), we obtain quenching decreases more or less inversely with the gas pressure. This means that the output power and the efficiency will have a maximum value that depends on both the discharge power (or total power) and gas (16) pressure. According to Fig. 5, the maximum efficiency before saturation as a function of discharge power is GIELKENS et al.: OPTIMIZATION OF THE MULTI-ATMOSPHERIC Ar–Xe LASER 257

Fig. 13. Here, the peak output power is plotted versus the gas temperature (a) for various pressures. Unlike Fig. 14, the drop of the peak output power is limited to about 30%.

and a subsequent decrease of the output by electron collision mixing according to (16). In Fig. 12(a) and (b), the laser output power density and electron density versus the input power density are plotted for three different gas temperatures at a pressure of 5 bar and an -beam current density of 0.9 A/cm . These graphs show that, although the increase of due to the temperature is relatively small, about 15% for a temperature difference of 100 K, the decrease of the output power can be dramatic at high input powers. In Fig. 13, the peak value of the output power for several pressures is shown as a function of the gas temperature. We see that when the temperature increases from 300–500 K the peak value drops about 30%. Because the heat capacity is proportional to the gas density, it is expected (b) that the higher the gas pressure is the larger the stationary Fig. 12. The dependence of (a) the output power density and (b) the electron period at a fixed power deposition, which is in agreement density on the discharge power density at various temperatures. Note the dramatic drop of the output power at high input powers, though the electron with Fig. 3. density increases only slightly. The temperature for which the laser may operate without degraded performance is also a strong function of the - attained at about 4 bar for both -beam current densities. beam current density, which is related to the dependence However, since the discharge pumping power at which of the electron energy on the -beam current as mentioned the output power is optimal is proportional to the pres- before. This is shown for a beam current density of 0.4 and sure (Fig. 4), one can expect that the maximum power 0.9 A/cm in Fig. 14(a) and (b), respectively. Here, the final may still increase with densities above 5 bar. gas temperature reached at the end of the stationary period, calculated from the dissipated discharge and beam energy, is plotted versus the total input power during that period. Thus VII. TEMPERATURE EFFECT ON THE KINETICS this figure shows to what extent the gas temperature may be From our analysis, it can also be deduced that the output increased at a certain input power level before degradation of is sensitive to the gas temperature because the rate constants the performance occurs. It is seen that the lower the electron of ion–electron recombination and the three-body collisions in density produced by the discharge, the less the increased the kinetic chain depend on the temperature. It is found that the temperature effects the quenching and consequently the output. higher the temperature, the higher the electron density so that At low input powers, however, the duration of the stationary the losses by quenching increase and the output power drops. period is merely limited by our pump duration of 20 s and the The rate constants of the three-body ion formation, and , indicated operating temperature is not the maximum allowable have a gas temperature dependence proportional to temperature that corresponds with that input power. whereas the dissociative rate constant has a temperature The maximum pulse energy defined as the energy during dependence proportional to where the stationary period depends not only on the gas pressure is the fundamental vibrational energy [12]. The temperature and power deposition, as we already described in the previous dependence of may be well approximated by because section, but also on the -beam current. According to Fig. 7, is only about 6 meV [13]. Substituting these dependences this maximum is about 10 and 16 J for -beam current into (7), we see for increasing temperature an increase of densities of 0.4 and 0.9 A/cm respectively. It should be noted 258 IEEE JOURNAL OF QUANTUM ELECTRONICS, VOL. 34, NO. 2, FEBRUARY 1998

(a) (b) Fig. 14. The calculated gas temperature, at the point where the laser output starts to drop, as a function of the total input power for several pressures. that if the total energy of a shot also includes the degraded part 10 J and 4% at 16 J . Depending on the repetition rate after the stationary period, the maximum total energy obtained and gas volume, high average output powers above 100 W in these experiments was 22 J at 5 bar. These results are are feasible. considerably higher than our previous results obtained with an -beam pulse of only about 1 s at the same conditions [7]. ACKNOWLEDGMENT Fig. 14 also shows that the increase of the -beam current density, which has a favorable effect on lowering the electron The authors wish to thank A. P. Napartovich and I. V. quenching, allows an increase of the gas temperature before Kochetov for their helpful discussions and calculations and degradation starts. For example, at a beam current density of N. N. Ustinovskii for his useful suggestions. 0.4 A/cm at 5 bar, the maximum temperature increase at the end of the stationary period is about 25 K for an input power of REFERENCES 25 MW whereas a density of 0.9 A/cm permits an increase [1] N. G. Basov, V. A. Danylichev, A. Y. Dudin, D. A. Zayarnyi, N. of more than 200 K at the same power deposition. N. Ustinovskii, I. V. Kholin, and A. Y. Chugunov, “Electron-beam- It is noted that in experiments with fission fragment pump- controlled atomic Xe infrared laser,” Sov. J. Quantum Electron., vol. ing Hebner [14] also found a dramatic drop of the output 14, pp. 1158–1167, 1984. [2] M. Ohwa, T. J. Moratz, and M. J. Kushner, “Excitation mechanisms power when the temperature increased beyond approximately of the electron-beam-pumped atomic xenon @Sd 3 TpA laser in Ar/Xe 400 K. In his experiments, the output power was insensitive mixtures,” J. Appl. Phys., vol. 66, pp. 5131–5145, 1989. to the gas temperature below 400 K. [3] A. Suda, B. L. Wexler, K. J. Riley, and B. J. Feldman, “Characteristics of the high-pressure Ar-Xe laser pumped by an electron beam and an electron-beam sustained discharge,” IEEE J. Quantum Electron., vol. 26, pp. 911–921, 1990. VIII. CONCLUSION [4] L. N. Litzenberger, D. W. Trainor, and M. W. McGeoch, “A 650 J e-beam-pumped atomic xenon laser,” IEEE J. Quantum Electron., vol. The experiments have revealed a stationary duration of the 26, pp. 1668–1675, 1990. output power that depends on the discharge power and gas [5] R. L. Watterson and J. H. Jacob, “Measurements of intrinsic efficiency pressure and not so much on the -beam current. After the and parameters of an electron beam pumped ArXe laser,” IEEE J. Quantum Electron., vol. 26, pp. 417–422, 1990. stationary region the efficiency drops. The output energy per [6] H. Botma, P. J. M. Peters, and W. J. Witteman, “Intrinsic efficiency and pulse for which the efficiency is at optimum is then equal to the critical power deposition in the e-beam sustained Ar:Xe laser,” Appl. output power times the stationary duration. These values are Phys. B, vol. 52, pp. 277–280, 1991. [7] H. Botma, P. J. M. Peters, and W. J. Witteman, “Saturation studies of plotted for different pressure in Fig. 7(a) and (b) as a function the e-beam sustained discharge atomic xenon laser,” IEEE J. Quantum of the discharge power. The efficiencies for the discharge and Electron., vol. 29, pp. 2519–2524, 1993. total input power are plotted in, respectively, Figs. 5 and 6. [8] Y. B. Udalov, P. J. M. Peters, M. B. Heeman-Ilieva, F. H. J. Ernst, V. N. Ochkin, and W. J. Witteman, “New continuous wave infrared We notice that the highest overall efficiency of about 8% Ar-Xe laser at intermediate gas pressures pumped by a transverse radio is obtained for low pressure and low discharge power. The frequency discharge,” Appl. Phys. Lett., vol. 63, pp. 721–722, 1993. [9] S. N. Tskhai, Y. B. Udalov, P. J. M. Peters, W. J. Witteman, and V. maximum discharge efficiency of 5%–6% is not so sensitive N. Ochkin, “Continuous wave near-infrared atomic Xe laser excited by to the input power, especially at high pressure. Looking for a radio frequency discharge in a slab geometry,” Appl. Phys. Lett., vol. the optimized output energy we find, according to Fig. 7, the 66, pp. 801–803, 1995. [10] S. W. A. Gielkens, P. J. M. Peters, W. J. Witteman, P. V. Borovikov, best result for a gas pressure of 4 bar and an input discharge A. V. Stepanov, V. N. Tskhai, M. A. Zavjalov, V. I. Gushenets, and power of about 8 MW . The output power will then have a N. N. Koval, “A long-pulse 300 keV gun with a plasma cathode for duration of 12 satan -beam current of 0.4 A/cm and 16 s high-pressure gas lasers,” Rev. Sci. Instrum., vol. 65, pp. 2449–2452, 1996. at an -beam current of 0.9 A/cm . The total efficiency and [11] J. J. Berger and S. M. Seltzer, Nucl. Sci. Ser. no. 10, NAS-NRC pub. output energies at these performances are, respectively, 5% at 113, 1964. GIELKENS et al.: OPTIMIZATION OF THE MULTI-ATMOSPHERIC Ar–Xe LASER 259

[12] T. J. Moratz, T. D. Saunders, and M. J. Kushner, “High-temperature V. N. Tskhai, photograph and biography not available at the time of kinetics in He and Ne buffered XeF lasers: The effect on absorption,” publication. Appl. Phys. Lett., vol. 54, pp. 102–104, 1989. [13] A. P. Hickman, D. L. Huestis, and R. P. Saxon, “Interatomic potentials for excited states of XeHe and XeAr,” J. Chem. Phys., vol. 96, pp. 2099–2113, 1992. [14] G. A. Hebner, “Gas temperature dependent output of the atomic argon P. J. M. Peters was born in Meerlo, The Nether- and xenon lasers,” IEEE J. Quantum Electron., vol. 31, 1626–1631, lands, on November 5, 1945. He received the M.Sc. 1995. degree from the Catholic University, Nijmegen, The Netherlands, and the Ph.D. degree in laser physics from the University of Twente, Enschede, The Netherlands, in 1981. He then joined the staff of the Quantum Elec- S. W. A. Gielkens was born in Heerlen, The Nether- tronics Group of the Department of Applied Physics lands, on May 2, 1969. He received the M.Sc. at the University of Twente where he is now an degree from the University of Utrecht, The Nether- Associate Professor. He has carried out research lands, in 1993. mainly in the ﬁeld of gas lasers. His Ph.D. work Since then, he has been working at the Quantum on a pulsed TEA CO laser was followed by research work on different types Electronics Group of the Department of Applied of excimer lasers, such as KrF, ArF, XeF(C->A and B->X), and on ionic Physics at the University of Twente, Enschede, The excimers. Currently, he is engaged in research on the vacuum ultraviolet Netherlands, on the high-pressure atomic Xe laser. molecular F2 laser and in rare-gas recombination lasers such as the atomic His interests include gas-laser excitation mecha- Xe laser emitting in the near infrared. nisms, laser kinetics, and laser spectroscopy.

W. J. Witteman was born in Monster, The Nether- lands, on December 12, 1933. He received the degree of mechanical engineer from the Technical University of Delft, The Netherlands, in 1958 and the Ph.D. degree in physics from the Technical Uni- versity of Eindhoven, Eindhoven, The Netherlands, in 1963. From 1958 to 1960, he was ﬁrst a Post-Doctoral Fellow and later a Research Associate at the Insti- tute of Fluid Dynamics and Applied Mathematics, University of Maryland, College Park, studying molecular relaxation phenomena by means of shock-tube experiments. From 1961 until 1969, he was with Philips Research Laboratories, Eindhoven, The Netherlands, where he was engaged with high-pressure physics and technology until 1963. After that he did research on water-vapor lasers, sealed-off COP lasers, and argon–ion lasers. Since 1969, he has been a professor at the University of Twente, Enschede, The Netherlands, where he works on high- power pulsed laser systems. He is actively engaged in the ﬁeld of COP lasers, mode-locking techniques, electron-beam and discharge-pulsed excimer lasers, both continuous and pulsed CO lasers, electro-ionization lasers like Ar–Xe waveguide lasers, and free-electron lasers of the Raman and Compton type. Since 1989, he has been a managing director of the Nederlands Centrum voor Laser Research (NCLR) B.V., which is a joint venture between the University of Twente and the industry. The NCLR develops advanced laser systems for industrial applications, such as a 1-kW Eureka-excimer laser operating at 1 kHz.